EAGER: SCH: Distributed and Adaptive Personalized Medicine

University Of Connecticut, Storrs CT

Investigators

Abstract

The quality of medical practice can be improved by the use of clinical data. However, current empirical approaches are developed on population averages and individual differences are often ignored. Furthermore, these approaches are static and fail to learn and change over time in an adaptive fashion. As a result, the science of medicine is not yet personalized and adaptive based on all the available data. Integrating medicine, engineering, and state-of-the-art information technology, this project aims at studying distributed and adaptive personalized medicine that collectively learns from an individual's clinical data in real time. This approach is inspired by collective and adaptive learning observed in nature (for example, honeybee swarming behavior, bird flight formation, etc.). The outcome of this project will enable smarter diagnosis, treatment, and prevention tailored towards individual patients. It will be disseminated widely through publications, seminars, workshops, and MOOC (Massive Open Online Course) development. The use of clinical data is at the core of medical practice today and, while various mathematical and computational approaches have been developed, conventional approaches are not geared towards individual patients or the dynamics of constantly changing clinical data. Inspired by studies of multicellular dynamics, this project explores distributed and adaptive personalized medicine which collectively learns from an individual's clinical data in real time through localized interactions. To make these efforts possible and scalable, this project will exploit a microservice (actor model)-enabled cloud cyberinfrastructure for increased accessibility, adaptability, interoperability, extensibility, scalability and sustainability. In addition, the result of this project, including the mathematical framework, can be applied to other domains, such as education, energy, telecommunications, and transportation. It will also be disseminated to academia through publications, seminars, workshops, and a MOOC to integrate the results of this work into interdisciplinary biomedical informatics research. All tools and documentation will be made available on GitHub so that a sustainable community can be formed around the project.